Hmm... actually, there's a bunch of tests like this already there, with comments that NaN errors are treated as zeros in this context.
Which suggests that I was wrong about what the fix should be. I'll shut up now... -- Raul On Fri, May 6, 2022 at 2:32 PM Raul Miller <rauldmil...@gmail.com> wrote: > > Looking at the J implementation, it's pretty clear that > jsignal(EVNAN); R 0; would signal the NaN error. > > But J has a macro -- NANTEST -- which I don't adequately understand, > and a variation of this logic is used in the NAN1 macro. > > (I was looking at the implementation to try and figure out what kind > of work is involved here... it looks like there's a number of routines > in va2.c and presumably va2ss.c which would need this test. But > finding the right place for the test would require understanding what > the test was testing...) > > -- > Raul > > On Fri, May 6, 2022 at 2:20 PM Raul Miller <rauldmil...@gmail.com> wrote: > > > > Yes: > > > > _. in an argument here should signal a NaN error. > > > > Thanks, > > > > -- > > Raul > > > > On Fri, May 6, 2022 at 7:15 AM Elijah Stone <elro...@elronnd.net> wrote: > > > > > > _. may give garbage results. It shouldn't hang. > > > > > > On Fri, 6 May 2022, Razetime wrote: > > > > > > > Although this does not seem like intended behaviour, it is worth > > > > mentioning > > > > that NuVoc <https://code.jsoftware.com/wiki/Vocabulary/underdot> > > > > explicitly > > > > states that _. is not recommended for use in J computations. > > > > > > > > On Fri, May 6, 2022 at 4:30 PM <jph.but...@mailoo.org> wrote: > > > > > > > >> Hello, > > > >> > > > >> > > > >> > > > >> _. does not mix well with logical operators. The following hang and > > > >> use a > > > >> lot of CPU on J 9.03 (Windows and Linux): > > > >> > > > >> > > > >> > > > >> 1 +. _. > > > >> > > > >> 1 *. _. > > > >> > > > >> +. / 8 0 _. > > > >> > > > >> > > > >> > > > >> Note, that I have no particular use for these sentences, I just > > > >> stumbled > > > >> across them due to a bug in my program. > > > >> > > > >> > > > >> > > > >> Philip > > > >> > > > >> > > > >> > > > >> > > > >> ---------------------------------------------------------------------- > > > >> For information about J forums see http://www.jsoftware.com/forums.htm > > > >> > > > > ---------------------------------------------------------------------- > > > > For information about J forums see http://www.jsoftware.com/forums.htm > > > ---------------------------------------------------------------------- > > > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
